Diseño y aplicación de una herramienta para la optimización de rutas de vehículos con aspectos medioambientales

Abstract

Nowadays, around 90% of city dwellers in the European Union are exposed to high concentrations of healthharmful pollutants, reducing the life expectancy of the population and having a large impact on the gross domestic product of European countries. Among the economic sectors, transport is presented as one of the main sources of pollution because it generates harmful levels of emissions and is responsible for up to 24% of greenhouse gases (GHG) emissions in the European Union. These emissions depend heavily on the fuel type used, the carried load, the engine technology and the total distance covered. The problem of the distribution of goods from warehouses to end users plays a central role in the logistics systems management, where the design of efficient routes is critical in reducing costs. This real-life problem is characterized by presenting a heterogeneous fleet where vehicles with different features are incorporated over the time for a better adaptation to the changing customer demands. These features include vehicles with different capacities and age, alternative fuels and motor technologies. Therefore, in the present context, environmental targets are to be added to traditional logistics strategies based on cost and time in the decision making process. The research of this Thesis has focused on the development of new mathematical models and algorithms for solving the Fixed Fleet Heterogeneous Vehicle Routing Problem with Time Windows (HVRPTW) with the additional consideration of reducing GHG and pollutants emissions. The formulation of the problem is made from two different perspectives. The first incorporates a methodology based on the estimation of the external costs of transport activities. The second perspective comprises a multiobjective optimization method with a priori articulation of preferences, in which the decision maker can establish the preferences in advance. The design of eco-efficient routes is proposed by linear mathematical programming models and is solved using quantitative techniques. These techniques include heuristics and metaheuristics that combine various advanced procedures to deal with the complexity of the problem. In particular, this Thesis describes a semi-parallel insertion heuristic and a hybrid variable neighborhood descent metaheuristic based on a tabu search algorithm for the local search and a holding list that achieves flexibility for solving any HVRPTW variant. The algorithms have been applied to benchmark problems from the scientific literature and to a real-world case that deals with a routing and scheduling problem in a service company with particular characteristics and constraints. The results show that the algorithm efficiently solves the problem addressed and it can be extended to other problem variants. The result of the Thesis is the development of a decision making process tool aimed to help in the design and control of eco-efficient routes. This tool can be integrated with the particular geographic information system (GIS) of each company, allowing the display of eco-efficient routes and assessing the economic, energy, operational and environmental impacts. Therefore, the tool will have an economic impact on the end users, with a comparison of the final routes and the results obtained from different alternatives, achieving greater competitiveness and fulfilling the sustainability commitments in the company. On the other hand, the tool globally contributes to a social improvement resulting from the fuel consumption and pollutant emissions reductions from road transport, which have an impact at local, national and international level. In this sense, the Thesis clearly contributes to the strategic development of the transport sector, increasing the efficiency of road transport fleets and achieving greater sustainability and competitiveness.